no code implementations • 21 Dec 2023 • Ken Trotti, Samuel A. Cruz Alegría, Alena Kopaničáková, Rolf Krause
We propose to train neural networks (NNs) using a novel variant of the ``Additively Preconditioned Trust-region Strategy'' (APTS).
1 code implementation • 30 Jun 2023 • Alena Kopaničáková, Hardik Kothari, George Em Karniadakis, Rolf Krause
We propose to enhance the training of physics-informed neural networks (PINNs).
no code implementations • 15 Jul 2021 • Alena Kopaničáková, Rolf Krause
We propose a globally convergent multilevel training method for deep residual networks (ResNets).
no code implementations • 28 Jun 2020 • Vanessa Braglia, Alena Kopaničáková, Rolf Krause
Our multilevel training method constructs a multilevel hierarchy by reducing the number of samples.
no code implementations • 13 Apr 2020 • Lisa Gaedke-Merzhäuser, Alena Kopaničáková, Rolf Krause
For our examples we employ a multilevel gradient-based methods.